The processes which initiate, maintain and terminate eating in disordered individuals have not been identified. We propose to evaluate the responses of subjects with anorexia nervosa, bulimia nervosa, or chronic dieting behavior to manipulations of energy density, food type, portion size, deprivation and food selection in naturalistic and laboratory settings before and after inpatient treatment. Four main hypotheses will be tested: First, that the eating behavior of eating behavior of eating disordered individuals differs quantitatively from that of the normal population on parameters such as rate of ingestion, dietary selection and onset of satiety. Second, that eating disordered subjects fails to adjust intake according to energy density relative to non- disordered subjects. Third, that the groups will show differential development of sensory-specific satiety and changes in rating of affective and cognitive variables to foods after test meal consumption. Finally, that the abnormal characteristics recorded in eating disordered patients will be normalized following treatment. These hypotheses will be tested in the following experiments which will systematically assess regulation, food selection and responses to foods: 1. Ability to regulate caloric intake will be investigated using a high and low-calorie preload or a liquid diet. The degree to which subjects compensate for caloric content will indicate sensitivity to internal regulatory cues. 2. Responses to familiar foods which are high or low- calorie will be studied. Food selection, amount consumed, rate of eating and ratings of sensory, affective and cognitive variables will provide evidence for disturbances of basic ingestive processes. 3. The influence of three levels of deprivation on eating behavior will be monitored. Adjustment of intake following deprivation will indicate responsiveness to internal signals of hunger. 4. Food selection and energy intake in a cafeteria and laboratory setting will be compared before and after treatment. The extent to which normalized levels of intake have been restored will be recorded. These experiments will permit us to identify parameters which characterize disordered eating. Such information will enhance not only our understanding of the pathogenesis of eating disorders but will increase our knowledge of the mechanisms underlying eating in normals.